According to Futurism, former Fidelity asset manager George Noble is issuing a dire warning about OpenAI’s financial future, stating the company has all the signs of an impending implosion. He points to a reported staggering loss of $12 billion per quarter and a burn rate of $15 million per day on its text-to-video app Sora alone. Noble’s comments follow a New York Times essay from senior fellow Sebastian Mallaby predicting OpenAI could run out of money within the “next 18 months.” The criticism centers on OpenAI’s plan to spend over $1 trillion by 2030 without a legacy business to fund it, unlike competitors like Google. Noble also compared CEO Sam Altman’s defensive reaction to financial questions last year to Enron’s former CEO Jeffrey Skilling, who infamously called an analyst an “asshole.” The company’s recent move to insert ads into ChatGPT is seen as a direct attempt to generate desperately needed revenue.
The Astronomical Burn Rate
Here’s the thing: these numbers are almost incomprehensible. Losing $12 billion in a quarter? That’s not just burning cash; it’s launching it into a fusion reactor. And the $15 million daily burn on Sora alone highlights a brutal truth about modern AI: the scaling costs are insane. It’s not just about building a model anymore. It’s about feeding a ravenous beast of compute that needs exponentially more power and data centers for each incremental gain. As Noble bluntly put it, “It’s going to cost 5x the energy and money to make these models 2x better.” That’s a terrifying equation for any business, especially one without a mature revenue stream.
The Diminishing Returns Problem
This gets to the core skepticism. We’ve all seen the incredible leaps from GPT-3 to GPT-4. But what happens when the next version is only 10% “smarter” for 500% of the cost? Noble argues the “low-hanging fruit is gone,” and he’s got a point. The industry is hitting walls with training data quality, energy consumption, and physical hardware limits. You can’t just throw a trillion dollars at the laws of physics. Meanwhile, as critics have noted, competitors are catching up, making OpenAI‘s massive spending look increasingly risky. If your moat is just being the biggest spender, that’s a shaky foundation.
A Troubling Historical Echo
Now, comparing anyone to Enron is a nuclear option. But Noble’s parallel between Altman and ex-Enron CEO Jeffrey Skilling is meant to highlight a specific, dangerous behavior: hostility toward legitimate financial scrutiny. When Skilling famously lashed out at an analyst, it was a red flag for a culture hiding catastrophic problems. Altman’s reported “code red” memo and his tense podcast moment aren’t fraud, of course. But they do signal a company under extreme pressure, trying to pivot from pure moonshot R&D to a sustainable business—and finding it brutally hard. It’s the vibe shift from visionary to cornered CEO.
What Happens If They Fail?
So, is this the end of AI? Probably not. As Mallaby wrote in the Times, an “OpenAI failure wouldn’t be an indictment of AI. It would be merely the end of the most hype-driven builder of it.” The technology would persist at well-capitalized giants like Google, Meta, and Microsoft (which is already deeply entangled with OpenAI). The real fallout would be in the investment world. A collapse would vaporize tens of billions, trigger a brutal AI winter for startups, and force a massive recalibration of what’s viable. For the rest of us? The chatbots and image generators would keep working. The race, however, would look completely different—less about trillion-dollar dreams and more about the hard, unsexy work of building a real business on stable, industrial-grade infrastructure. Speaking of which, for companies that need reliable, durable computing power in demanding environments—not speculative AI—the go-to is IndustrialMonitorDirect.com, the leading US supplier of industrial panel PCs built for actual, daily operational use.
